LAS dataset scalability

The LAS dataset is scalable, allowing you to work with large point cloud datasets quickly and efficiently. Rendering very large point sets or TIN-based surfaces is generally quite difficult, partially due to limitations of hardware. LAS datasets are beneficial in that you render the areas that you need to visualize at an optimized point resolution.

For optimal results, ensure that the LAS files to be used in the LAS dataset are all reasonably sized. A recommended file size is approximately 500 MB when statistics are calculated and able to provide LAS dataset auxiliary files (.lasx). A spatial index is calculated and added to the LAS auxiliary files when statistics are calculated. If statistics are not calculated, then a recommended file size is approximately 200 MB or less. It is recommended that statistics be generated to provide the auxiliary files which contain a spatial index.

On-the-fly surface construction, display, and analysis are faster for smaller-scale applications because only a thinned subset of the data is required. LAS dataset tools facilitate the use of large point collections, such as lidar, that would normally pose display problems. Scalability is achieved through random point thinning. The original data is not moved or averaged in any way.

The LAS dataset is designed to be used with LAS files that have originated from airborne lidar data sources. However, LAS files created from terrestrial lidar data can be displayed and analyzed in ArcGIS as point clouds.

Scalability in ArcMap

In ArcMap, points are thinned based on the current extent and the number of points estimated to be within that extent. ArcMap reloads the data every time you zoom or move throughout the display. Each time the display is panned, only the data necessary for display within the current extent is accessed and displayed on-screen. The LAS points are thinned randomly based on the extent and number of potential points to display. The thinning is the same in ArcMap for either surface or point display of a LAS dataset. The surface view of the LAS dataset is triangulated based on the source lidar points.

If a LAS dataset is created from many LAS files, the minimum bounding boxes representing the extent of each LAS file will be displayed initially when the LAS dataset layer is loaded in ArcMap. If there are fewer lidar points in the referenced LAS files than the overall point budget, all the lidar points will be displayed. The initial display will be a thinned point set. As you zoom in, more points will be added to the display.

The LAS dataset layer indicates a percentage in the table of contents describing how much the data in the display window has been thinned. As you move throughout the LAS dataset, you will see Data percentage change accordingly. Data percentage is a value indicating the percentage of points used to display the current LAS dataset relative to the entire number of points. This point percentage is based on the total number of points in the display extent, not on any filters that have been applied to the current LAS dataset. The graphic below shows the legend of a LAS dataset displayed using an elevation renderer with 100 percent of the points. This indicates that the lidar data is being displayed at full resolution.

LAS dataset layer in the table of contents

Setting full resolution in ArcMap

When working with LAS datasets that contain more points than can be viewed at one time, you can either view a randomly sampled subset of points or zoom in to an area to see full-resolution data.

A full-resolution point set means that every point is displayed and none of the points are thinned away. Often, lidar data is oversampled for many applications, and a thinned point set relative to full resolution is acceptable for display and analysis. However, for other applications, such as data validation and investigation, it is common to request the data at full resolution. To obtain a full-resolution point cloud in ArcMap, either zoom in to the data until the LAS dataset layer indicates a data percentage of 100 percent, or choose Zoom To Full Resolution from the LAS dataset layer's context menu, or use the Full Resolution Scale option from the Display tab on the LAS dataset Layer Properties dialog box.

Changing the default rendering preferences

There are options available to allow you to manipulate the default point rendering for a LAS dataset.

Advanced access and options can be found on the Display tab of the LAS dataset Layer Properties dialog box.

The default rendering preferences are quite conservative and can be modified for more powerful machines. Most machines should be able to handle a surface at the default resolution; however, it is possible for some basic machines to find this setting too ambitious. The rendering performance can vary significantly from machine to machine, and you can choose to override the default setting based on how your machine is performing.

LAS dataset layer properties

The Display tab on the LAS dataset Layer Properties dialog box provides the following rendering parameters that control scalability: a Point limit property, a Point Density slider bar, and a Full Resolution Scale factor.

LAS dataset Layer Properties dialog box

The default point density is set to 800,000, which is fairly conservative. This means that unless the number of points is fewer than the budgeted limit, the display shows fewer than 800,000 points. Increasing the limit will allow more points to be used if needed. Given that this limit can vary significantly from machine to machine (because of hardware, operating system, and other such factors), you can choose to override the default setting based on how your machine is performing.

To increase the point limit for powerful machines or decrease it for less powerful machines, follow these steps:

  1. Right-click the LAS dataset layer in the table of contents and click Properties.
  2. Click the Display tab.
  3. Change the Point limit property as required.
  4. Click OK.
NoteNote:

The maximum number that can be set for Point limit is 5,000,000.

The Point Density slider bar allows you to increase or decrease the number of points thinned by the default thinning algorithms in ArcMap. The Points per centimeter on screen value will change as the slider bar position is moved right or left.

When Use scale to control full resolution is checked, a scale factor can be set in the Full Resolution Scale window. This is a scale factor used to control when the LAS dataset will render without thinning, using 100 percent of the LAS points. It is used when the map scale is equal to or greater than the scale you specify. The point limit is still honored, though, so if the number of estimated points for the current extent exceeds the limit, the LAS dataset will thin and not be drawn using all the data. When this occurs, an asterisk is displayed next to the data percentage listed for the layer in the table of contents. When the map display scale is less than the full resolution, scale thinning will occur based on the setting of the Point Density slider bar.

Scalability in ArcScene

ArcScene is a memory-based application and therefore has an upper limit to the complexity of the surface or number of lidar points that it can display. ArcScene will load only the applicable points when the layer is first loaded or reloaded when certain layer properties are changed. This means that unless the total number of points in the LAS dataset is less than the budgeted limit, the surface view in ArcScene is at a static level of detail that does not include all the points.

In many situations, LAS datasets contain more points than ArcScene can handle; therefore, ArcScene will load only a thinned point set into the scene. It will draw points from every referenced LAS file of the LAS dataset, randomly selecting points from each file equally; therefore, an even number of points will be displayed initially from each LAS file if the LAS dataset is quite large.

The default point limit is 12,800,000 for LAS datasets displayed as points, while the default point limit for LAS datasets displayed as surfaces is 800,000. Both of these values are quite conservative and may be increased for powerful machines.

TipTip:

While most machines should be able to handle a surface or point cloud at the default resolution, it is possible for some machines to find this setting too ambitious. Also, it is worth noting that the surface point resolution is defined for each LAS dataset layer separately, so be sure to make allowances when displaying multiple LAS datasets; that is, if you are displaying five LAS dataset layers as surfaces, each with a resolution of 800,000 points, there would be a total of four million points used for surfaces in the current view.

Given that this limit can vary significantly from machine to machine (because of hardware, operating system, and other such factors), you can choose to override the default setting based on how your machine is performing.

To increase the point limit for powerful machines or decrease it for less powerful machines, do the following:

  1. Right-click the LAS dataset layer in the table of contents and click Properties.
  2. Click the Display tab.
  3. Change the Point limit property as required.
  4. Click OK.

Setting full resolution in ArcScene

When working with LAS datasets that contain more points than can be included in a surface at one time, you can either view a randomly sampled subset of points or define an area of interest within which to see the lidar surface at full resolution.

A full-resolution point set means that every point is displayed and none of the points are thinned away. Often, lidar data is oversampled for many applications, and a thinned point set relative to full resolution is acceptable for display and analysis. However, for other applications, such as data validation and investigation, it is common to request the data at full resolution. To obtain full resolution in ArcScene, you can set an area of interest. For more information on how to set an area of interest in ArcScene, see Displaying LAS datasets in ArcScene using an area of interest.

Related Topics

3/5/2014